Awesome

SWSL

We provide the implementation of Semi-Weakly-Supervised Learning of Complex Actions from Instructional Task Videos based on MuCon.

Preparation

Please follow the instructions in MuCon for environment installation and data preparation.

Download the files of video lists for different ratios of weakly-labeled videos and put them in the dataset folder. TODO: update download link.

Running

Please run this command to train and test the model:

 python src/train_test_swsl.py --cfg src/configs/docker/inside.yaml --set add_dataset.ratio 0.1  --set dataset.split 1

You may change the ratio of weakly-labeled videos and dataset split. You may also change other parameters defined in src/configs/mucon/default.py.

Citation

If you find the project helpful, we would appreciate if you cite the work:

@article{Shen-SWSL:CVPR22,  
         author = {Y.~Shen and E.~Elhamifar},  
         title = {Semi-Weakly-Supervised Learning of Complex Actions from Instructional Task Videos},  
         journal = {{IEEE} Conference on Computer Vision and Pattern Recognition},  
         year = {2022}}

Contact

shen [dot] yuh [at] northeastern [dot] edu

Acknowledgement

The code-base is built using the fandak library and MuCon repo.